Filename column for multiple file read

wahi80 · Post by **wahi80** » Thu Jan 24, 2013 11:14 am

Hi,

I have a job where I read multiple files using the File pattern approach.
I also needed to get the file name , so I used the filename column option.

My file pattern is ??.txt.

While my job read all the data in the files correctly, the name in the filename column is always ??.txt. I was expecting to get 1a.txt, 2b.txt and so on.

Is this the way it is designed to work?

prasson_ibm · Post by **prasson_ibm** » Thu Jan 24, 2013 11:26 am

In case of file pattern,datastage perform just like unix "cat" command,since all the metadata of files are same,so its read bunches of the files togather.

wahi80 · Post by **wahi80** » Thu Jan 24, 2013 11:45 am

But that is not my requirement, I need to map filename to each row

chetan.c · Post by **chetan.c** » Thu Jan 24, 2013 12:47 pm

Set this environment variable APT_IMPORT_PATTERN_USES_FILESET to true and check.

wahi80 · Post by **wahi80** » Thu Jan 24, 2013 12:50 pm

That works!!!

chulett · Post by **chulett** » Thu Jan 24, 2013 1:56 pm

FYI - a search here would have turned that information up.

chetan.c · Post by **chetan.c** » Fri Jan 25, 2013 12:39 pm

Apologies Craig.
I Should have pointed towards it.

chulett · Post by **chulett** » Fri Jan 25, 2013 1:54 pm

No worries! Just wanted to add that little tidbit.